Mentor of a team that runs the KOP drive train and we are discussing the switch to swerve. We use the rev ecosystem and if we did move to swerve I am intrigued by their modules that are built for NEOs and 550s. Would love to get some feedback from anyone that has competed with it and if you liked it or wished you had gone with the modules from SDS or some other vendor.
Some recent posts I have made on the topic:
4 Likes
I inspected quite a few robots with the Rev swerve modules this year, and other than some tread delamination issues, I heard nothing but praise for them.
8 Likes
The modules are amazing. Can’t recommend them strongly enough. Coming from a team (6045) that historically hasn’t run anything other than the KOP drivetrain… it worked really well for us, and was a huge contributing factor for our successful season.
8 Likes
We used Rev swerve this year (our first year running swerve!) and had an overall great experience! These modules work really well with the rest of the rev ecosystem and using maxtube in your chassis construction is super convenient since all of the bolt holes already line up. These are also very easy to assemble, since we found them simpler to assemble/service than our WCP SwerveX. The only thing to be aware of with these modules is that you will likely have to re-tread the wheels since the tread that comes in the kits can slide off the wheel and wears down pretty quickly. Other than that, I strongly recommend these modules!
3 Likes
Rev swerve were awesome for us as well and key to our success. It was so fun to drive and quick to set up with code and mechanical. We were planning on running west coast, but had the swerve running in a week during season after they arrived. We are buying more for next year, but could still use the original set if we broke down the robot for parts.
tips-
- Aluminum wheel hubs with black neoprene lasts a whole regional. We did not have as much luck with the tread versions. Other teams did, so it might have been our gameplay.
- Superglue the little motor key in the slot.
- 3d print the spectrum module covers to minimize metal finding its way into the modules.
- Rev structure makes assembly of frame super easy. Take care to use a square and careful sets of hands when assembling.
- A spare set of wheels on hand for issues that arise, including bevel gear(part of wheel assembly), makes for sub 5 minutes changes. It was as fast as changing our bumpers.
We love the modules.
7 Likes
For newcomers this is the ideal swerve module simply due to the code that comes with it, cost, along with all the gearsets included incase you want to change speeds.
Only thing I highly recommend is getting the 3"x1" aluminum wheels and 1" wide tread from The Thriftybot.
If you feel your programmers are pretty solid at working through little bugs and issues, then I highly recommend Swerve Drive Specialties or the West Coast Products SwerveX purely because I feel the larger wheels will be more usable in future games.
We run SDS mk4i and loved them.
7 Likes
We ran the MaxSwerve Modules for a total of ~80 matches this season across (4) events, here are some notes.
- As mentioned above, the original (V1?) wheels aren’t the greatest. We ran these wheels through our first event and were able to get away with ~1 wheel swap per 8-10 matches, but we also ran the REV Slew Rate Limiters with some slight modifications.
- For Events 2, 3 and 4 we ran the TTB Replacement Wheels (1" Wide Aluminum) with Black Roughtop Tread and had no issues with traction or tread life. If you do run the TTB Setup, you’ll need to either make a custom Zero-ing Fixture or Zero without the wheel mounted, neither is that big of a deal.
- Modules may be a bit slower than some of the “Fast” Swerve Drives - we originally ran the 12T Pinion, Later Swapping to 13T. In general, we could cycle in the “Above Average” Range (Something like 8 cycles average per match) but we were hard limited by speed at some level. (We were also 120ish lbs, so lighter robot could likely run the 14T ratio for a bit more speed.)
We did encounter a unique failure mode at the Championship - on (2) different Occasions, the Front Left Module Lower Plate Failed at the Azimuth Bearing “Step” / Seat / Retaining Feature.
As designed, this feature is machined into the bottom plate at a material thickness of .030" - essentially it’s a step to keep the bearing from moving “Up” into the Lower Plate - 10-32 BHCS keep the bearing pushed against that step / keep the bearing from falling out. One of these failures was in Milstein Qualifications and quickly fixed via a parts swap, the second failure happened during Milstein Final 2 and benched us for F3.
Considering how unique that failure appears to have been (I’ve yet to hear about anyone else experiencing the same thing) it may have been hyper local to our implementation and/or due some unique feature of the Field on Milstein. Anecdotally, all signs point to the failure happening while traversing the Charging Station, there was a spot that appeared “Soft” that may have exposed an edge when cresting the platform - if the wheel caught there just right, it may have seen some atypical extreme load that would have caused the failure mode we saw. The Aluminum wheels may have also played a factor, as they would not have dampened any sort of lateral wheel contact in the same way the Overmolded Wheels would have, though I suspect in those cases, it’s likely that the wheel itself may have delaminated, which is arguably the less expensive failure mode.
The jury is still out as to whether or not we will use the modules in upcoming seasons. If the design for the Retaining Step is revised in the lower plate, then it’s possible - but otherwise we’ll likely be moving to SDS or WCP.
That said, ignoring the failure mode we experienced at the CMP, the Modules were generally pretty easy to live with all season - more often than not, they just worked - and when they didn’t, it was almost always related to wheels or general maintenance type things. Prior to the CMP we never had a module failure during a match, so there is a way to compete with them successfully, they just may not be robust enough for all versions of gameplay.**
One last note, the modules tend to run on the loud side. This seems to be especially true when running the “Slower” pinions as they appear to be an Addendum Shifted Profile and the Teeth are less Perfect. I believe the steering input gear is also Shifted as well, though the “Whirr” it makes is a bit less annoying.
**It’s worth noting that the initial failure happened on a Module that had previously seen at least 60 or so matches of gameplay, while the second failure happened with less than 10. This points to the failure being due to a very specific load case and not something like fatigue or some other repeated abuse.
3 Likes
Do you have any pictures of the failure?
First Image shows the (2) failed modules (note how the bearing sticks out at the same place), Second Image shows the Flange Partially Sheared from the Main Plate. Referencing the CAD, Spec on this is ~.031" Thick with a sharp corner. You’ll notice small tabs in the second picture - there’s also a section of extremely thin material right where the bearing retainer tapped holes are, it’s something like .025" thick between the Major Dia. of the thread and the “Edge” of the Step, this may also have something to do with the failure.
Also ran a quick simulation (sorry no result shot handy), but basically if you apply a force laterally and up into the wheel, this failure mode generally makes sense. (Though it requires more than the weight of a robot, so it’s likely caused by the “Right” impact)
4 Likes
Thank you for these, I new there would be some threads out there on this but couldn’t find them.
1 Like
Thanks to everyone for the feed back. This was exactly what I was looking for.
Its interesting that you say the modules are loud. In comparison to our sds mk2 modules the rev swerve is significantly quieter. The sds mk2 modules were so loud that when driving in the shop you could barely hear someone talking to you.
I think the input pinion that’s installed makes a big difference.
We started with a 12T Pinion and the modules were loud enough that most people didn’t try to talk when we drove around our lab. 13T Pinion ended up being significantly quieter, I’d imagine that the 14T (Assuming the 14T is the non adjusted Pinion) is quieter still.
Anecdotally, we’ve noticed the same thing with the VP 10T, 11T and 12T Pinions. Seems to just be the nature of modifying the addendum?
My point of reference in this case is compared to (Any?) WCP Module / SDS MK4i, both of those have the standard “Spur Gear Whirr” type noise.
I don’t know if anyone has said this, but REV has new wheels up for sale.
I beleive that these will let significantly longer and will perform great.
Edit: looks like they don’t.
2 Likes
From my understanding 3005 was running these at champs and had delamination a couple times in playoffs.
Someone from their team can confirm or correct me, but that is my understanding. I would personally use the TTB 1in wide aluminum hubs with tread.
1 Like
6045 was swapping them out every match. We tried going every three matches at first, but this led to issue in two qualifications matches at champs. We actually experienced less issues with the V1.0 wheels. Ultimately, the wheels caused failures in two early matches for us at champs. Probably played a part as to why we fell to the 10th pick.
I will be the first to admit that the wheel problem is not solved yet. While the 1.1 wheels worked better and have better wearing than the 1.0, high speed impacts on the field can still cause de-laminations.
We have a few pathways for improvement and the tooling for the wheels are going to be undergoing some work so we can change to some different materials on the core.
At this moment in time, I would say, if you just want to try swerve you will be fine with the molded wheels as they are. If you you want to take this to an off season, swap to the metal wheels and run tread like all the other modules on the market.
26 Likes
My team purchased six REV MaxSwerve modules with the “fast” ratio and put them on the second of three robots we built during this game’s build season. Initially we intended to run this robot in competition week six, but after our week two and four events (in St. Louis and Colorado respectively) we decided that we had run out of time and money to continue trying to maintain and develop multiple robots that drove and performed differently to each other, so we reluctantly let that robot sit for the rest of the season, and continued to compete with our SDS MK4i modules. We did drive it for our local week zero event and in practice at our facility. After about 10 matches worth of field time we had a wheel delamination and after about two hours of drive time one of the modules got noticeably noisy, but at that point we benched it so still need to investigate.
Planning to continue using the in the future for our summer and fall off-season build. We will likely purchase the aluminum wheels, or wait until the molded wheel issue is resolved by REV.
As a REV sponsored team, we kept track of robots at all four events we attended that ran this swerve design, and we brought spare wheels for teams at the three Regionals, in case the delamination issue came up (we did not bring these to the championship because we knew REV would be there). Those wheels were never used for delamination replacement but just for normal wear.
In self-interest, we also kept track of all the types of drive bases run by teams at our events, to make sure that we could help our partners play their best, and to help us pick reliable partners for elims. The three brands of module that tied for our highest serviceability and reliability were SDS, WCP, and REV, and notably these types of drive base were generally well driven and good contributors to an alliance as well.
On the software side, it took our programmers significantly less time to get the REV modules running well than the SDS modules, mostly because they had to re-start their SDS coding with the version team 364 developed, while REV provided usable code “out of the box”. This wasn’t a significant issue for us (since we had a year’s head start) but as I visited with teams in St. Louis and Tulsa this year who are thinking of trying swerve for the first time, I thought it was a good feature to consider.
One thing more, I do think that REV’s vertical integration of this product with the rest of their line is worth considering. It’s a pretty unique way of approaching the FRC parts market at the moment, rather than creating ad hoc parts with discrete functions. Of course there are risks for a team “all-in” on REV stuff but there are advantages too.
8 Likes
We used REV swerve this year which was one of the factors enabling us to have our most competitive season yet.
Relative to cost its fantastic, and is suuuuuper small which is really nice to design around.
The molded wheels were an issue but we never had issues with the billet+tread. We did have to run our 2.5 auto instead of 3 piece auto in hopper elims partly because we switched from 1.1’s to the billet wheels for elims to avoid the frequent changeouts. We switched out plastic wheels every match to every other match depending on condition. We definitely had a few matches where the robot was a lot less controllable after heavy hits during quals.
We experienced less delamination running the 1.0 wheels earlier in the season, though at the time our robot was under 70lb.
I can’t say enough positive about how easy it was to assemble and program using REV libraries. I think it is a great go-to module for teams getting ready to switch to swerve without a lot of controls implementation experience.
1 Like